: SMAD proteins are intracellular signaling molecules of TGF-beta and have been shown to play a pivotal role in gastrointestinal carcinogenesis. The applicant's group has recently reported a novel phenotype with heterozygous Smad 2/3 knockout intercrosses, with all mutants dying at E14 with marked liver hypoplasia and holoprosencephaly. The Smad 2/3 mutants were notable for a marked fall in the expression of ELF 3, a beta-spectrin previously cloned by the applicant's group. Antisense studies suggest a crucial role for ELF 3 in biliary epithelial cell formation. In addition, Smad 2 and Smad 3 bind to ELF3. Experiments in this proposal are aimed at testing the hypothesis that ELF 3 plays a pivotal role in TGF-beta signaling involving both gastrointestinal and anterior cranial development. The specific aims are: 1. (i) To determine the phenotypic and molecular basis of the size heterogeneity of ELF proteins, and (ii) to test whether ELF can associate with other signal transducing proteins to gain an understanding of the mechanism of action of ELF. 2. (i) To characterize the interaction of ELF with Smads and to define the domains in Smads that bind ELF, (ii) to confirm interaction of ELF and Smads by co-immunoprecipitation experiments, and (iii) to investigate whether TGF-beta receptor-mediated phosphorylation of Smads or ELF may regulate their interaction. 3. To study the functional importance of ELF in TGF-beta signaling by defining the domains in ELF that interact with Smad 2, its subcellular localization, and its association with the TGF-beta receptor. 4. To determine whether ELF regulates the subcellular localization of Smad 2 and Smad 3 based on the preliminary evidence suggesting that ELF functions upstream of the Smad signaling pathway. These studies should lead to a greater understanding of the specific role Smad 2, Smad 3, and cytoskeletal proteins such as ELF in both gastrointestinal and anterior cranial development.